AndrewKemendo

Kind of a weird argument. I’m the only one in the thread actually talking real dollars and the response is “that doesn’t last forever?”

When will that be?

I just finished a $23k engineering job in 3 months, entirely solo, that would have taken a year without these tools. That’s real money to feed my kids and working tech that made a client happy.

That’s one job and I have several in the pipe.

The “financial payoff” for no more than a few 100 a month at absolute most is there and it’s real. If you aren’t seeing it then you’re doing it wrong.

I usually don’t even spend more than $20 a month each for chpt and claude, deepseek free is solid. I am about to switch fully local too so that should drop to basically cost of watts.

I dunno what yall are doing

Agreed. After a long time in tech the first movers almost always get killed

In fact most people aren’t even aware of most first movers cause they shut down or get acquired and their tech get absorbed into some giant conglomerate

“More done at once”

Why do you need to get more done at once?

It sounds like you’re describing a one shot but with more steps

I stopped thinking about Apple as a technology company when I realized they ACTIVELY fast follow instead of being pioneers in fundamental technology.

I don’t think that’s an insult or bad, if anything it makes them more reliable as a company because they ride the wave of technological advancement rather than on the leading edge.

That’s a really great position as long as you have no ego about being the first - and they repeat over and and over that are THE BEST not the first.

Being third for 100 years beats being first for 40 years and Apple has the Middle Kingdom strategy

How are people hitting limits?

I consider myself a Heavy cgpt user since 2022. To the extent where I’ve made over 30k ish in solo dev side-projects since 2024 that wouldn’t have been possible without hiring a team.

I have not once gotten a over limit notice from openai.

Alternatively, I used to max out my claude tokens very quickly even on fairly small projects.

The only difference I can tell from how I see other people utilizing these tools is that I never once “fire and forget.”

My project flow is: 1000 rounds of text back-and-forth on architecture, data structures, competitive analysis, tool use and calling, etc… and then the last 10% is code writing test development and then usually about five rounds of CI/CD iteration until I get to a MVP or something shippable.

Are the majority of people attempting to one shot and then when they fail they just keep trying to one shot?

In my experience, you can get 90% of the way to the technical solution before generating any code whatsoever.

That’s been true before LLM’s continues to be true now.

He gave his talk first on day 2 of the AGI 26 conference

https://m.youtube.com/watch?v=kQBwtgkHnC4&t=289s&pp=2AGhApAC...

I mean Roger Penrose proposes that consciousness is inside the waveform collapse so…

Many definitions of intelligence and ultimately consciousness rely on fanciful ideas and complex theories

Most human societies claim dominance over “everything else” aka “nature” and in order to do that you need a reason “humans” are different from “nature”

Is this supposed to be computer generated satire?

Edit: Ah, yes:

> Dsrupt is satire about technology's takeover of everyday life, written with AI's help as it happens. The trends are real. The people and companies are made up. A human edits, reviews, and signs off on every story.

Living in a machine Oroborous

Send it to [email protected] and I’ll respond from that

Reach out if you need moral support at some point.

I’ve been saying we need a general strike for a long time and I’m hoping it will happen

https://generalstrikeus.com/

> “It has to be a deeply held ethical commitment that’s analogous to a religious commitment,” he said. “That’s a fact question that presumably the employer will investigate.”

I’m curious if the opposite would work:

“As a transhumanist my religion compels me to share everything with my machines, otherwise they can’t ensure my mind uploading is up to date “

I see no reason transhumanism is any less of a religion than humanism.

I'm not sure it's possible to even give a reliability score for a new product.

The only way to do this would be to give reviewers the product a year before it's released. That's obviously not possible in the modern JIT product market economy

The reason I only buy old stuff and basically never a new product (unless it's a early adopter type thing) is so that I can get the community reliability score.

Never pay to be a new product tester

Means you’re out of money

Having been either CTO or CEO or Director/Principal a few times these are the kind of phrases that should always trigger you to immediately find new work:

“Like a family” “Sacrifices” “New platform” “Come together” “Critical delivery” “Recap”

I learned this from another senior when I was at Unity in 2022 that dipped immediately when they started talking like that.

Exactly. If the function is already verified to be correct, then the unit and integration tests will determine if the function is appropriate for the context.

Unit tests aren't for verifying functions.

Yup

I’m running local models more and more and starting to move to just tailscaling into everything

What if i told you

User testing is part of test coverage

Indeed ;)

It’s a question of the degree to which you want a self audit to be native to your structure or not

if you are confident that the complexity of what you have built can be fully tracked in your head then either you are some sort of God or you’re not used to building complicated enough things with critical failures attached to them

Again though why have midterms or finals?

Break down the learning objectives to measurable experiment replication then:

Show one, do one, teach one

If you can’t do one or teach one you dont move to the next step.

It’s literally just how apprenticeship works and there’s no reason you can’t do that for calculus or intro to economics and especially applicable to EE, ME, CE

Teaching in academic settings is absurdly inefficient and counterproductive in my experience

I’ve probably worked on 100+ software and hardware products for consumers, military and industry

.0001% of that software had formal verification, bitwise memory management, 100% test coverage, clean documentation and specifications, a clear roadmap, long term sustainability plan and sustainable economics

Technology is like a good meal: only possible for a moment and only a moment.

too early and it’s inedible too late and it’s spoiled.

I never saw any point in grading work ever

I went to Montessori school as a kid then I was in private school briefly till my parents went broke. Then I went to public school for the remainder of my career before I went to college and graduate school. I’ve taught at the graduate and undergraduate level and sat on academic boards for Purdue.

Graded and ranked assignments do nothing other than act as a filter for the structural organization to determine who they want to give resources to

The “educational” system has nothing to do with platonic or frankly any other thoughts of education for the polis

Good riddance

That’s if I was going to build my own model on myself, which I’m not planning on doing at this point

I have an insane amount of writing output as training data going back a loooong time so it all comes down to how much training data do you have

Couple things:

I have a “style guide” markdown that is pretty expansive

I run an RLHF style loop with me in the middle for all my writing now where I give a voice rant that is over expressive (aka redundant samples of concepts that I want to express clearly) in order to manage lossy compression steps

Then I will manually clean up that last 5% after a half dozen rounds

That sequence goes into the style guide to be referenced as a reasoning chain of thought to map rant > output

It’s been three years of tuning I’d suspect at this point

I wouldn’t be so sure that that’s human text

I can reliably get my custom models to output a 95% perfect replication of my tone and voice

Yeah I’m not sure what happened but there was a huge bot swarm the last few days

I had an open gitea host pulled into some botnet

Luckily digitaloceam sent an alert.

Who said it’s new? :)

However if you look at the trust busting era what preceded it was this exact kind of private “cooperation” to prevent government involvement